7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

1/90

TRIBOLOGICAL BEHAVIOR OF NANOCRYSTALLINE METALSPREPARED BY SURFACE MECHANICAL

ATTRITION TREATMENT

by

MICHAEL FRINK

Presented to the Faculty of the Graduate School of

The University of Texas at Arlington in Partial Fulfillment

of the Requirements

for the degree of

MASTER OF SCIENCE IN MATERIALS SCIENCE AND ENGINEERING

THE UNIVERSITY OF TEXAS AT ARLINGTON

August 2009

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

2/90

Copyright by Michael Frink 2009

All Rights Reserved

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

3/90

ACKNOWLEDGEMENTSI would like to thank my future wife, Donna Chen, for her overwhelming support while I

was working on my degree. Her encouragement kept me focused and on track to finish my

degree. I would also like to show my great appreciation to my research advisor, Dr. Efstathios I.

Meletis, for his advice and guidance throughout my research and studies. I extend my

appreciation to my committee members Dr. Pranesh Aswath and Dr. Michael Jin for their time,

effort, and patience. I acknowledge the use of the UTA Characterization Center for Materials

and Biology (C2MB). Without this facility, analysis of my materials would have been much more

difficult. I also would like to thank Punnapob Punnakitikashem for helping me to develop the

thermally evaporated aluminum coating at the UTA Nanotechnology Research & Teaching

Facility (NanoFAB). Finally, great thanks goes to my fellow group members for all of their help

and support along the way, especially Dr. Maria Hossou, Cristian Cionea, and Razieh

Khalifehzadeh.

May 22, 2009

iii

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

4/90

ABSTRACT

TRIBOLOGICAL BEHAVIOR OF NANOCRYSTALLINE METALS

PREPARED BY SURFACE MECHANICAL

ATTRITION TREATMENT

Michael Frink, M.S.

The University of Texas at Arlington, 2009

Supervising Professor: Efstathios I. Meletis

Nanocrystalline materials have shown great promise for a number of engineering

applications ranging from structural and electronics to biomaterials and drug delivery.

Nanoscale grain structures can cause tremendous improvements in strength, hardness and

other mechanical properties. This is partially due to the suppression of dislocations within

nanocrystalline grains and the entirely different deformation modes active at that scale. In spite

of anticipated benefits, very little work has been performed to explore nanoscale effects on the

tribological behavior of materials. Surface Mechanical Attrition Treatment (SMAT) is an

effective and economical method to produce nanostructured surface layers in bulk materials.

The process only takes hours to apply and industry is familiar with the similar process of shot

peening. In the present work, SMAT was utilized to develop a gradient grain size

nanocrystalline layer in pure nickel, aluminum, and titanium. The SMAT microstructures

produced were characterized by surface profilometry, scanning electron microscopy in

conjunction with energy-dispersive spectroscopy, x-ray diffraction, and nanoindentation. Pin-

iv

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

5/90

v

on-disk wear tests were performed on the SMAT processed materials to study the effect of the

produced gradient microstructure on the tribological response with comparison to their

microcrystalline counterparts. Wear testing showed that in general, the nanostructured layer

reduces the coefficient of friction and improves the wear resistance but the extent of the

improvements is material dependent. The largest reduction in friction was observed for pure Ni

(~33%) followed by pure Ti (~14%). Minor effects on friction if any (< 1%) were observed for

pure Al since its behavior was dominated by aluminum oxide formation at the asperity contacts.

The SMAT treated Ni and Ti display friction characteristics similar to those of nanocrystalline

materials possessing a uniform grain size. However, surface roughness resulting from the

SMAT process may have contributed to somewhat initial higher wear rate. In both of the above

cases, finer wear debris was produced for the SMAT processed metals suggesting a different

wear mechanism compared to their microcrystalline counterparts. SMAT aluminum exhibited a

minor reduction in friction compared to its microcrystalline counterpart and an earlier stage for

aluminum oxide debris formation during wear. The faster aluminum oxide debris formation

could be due to surface roughness effects, instead of the applied SMAT layer. The produced

SMAT layer was not thick enough to have any appreciable effect. The present study shows that

SMAT processing possesses good potential as a surface nanocrystallization/modification

method to improve tribological response of certain engineering materials.

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

6/90

TABLE OF CONTENTS

ACKNOWLEDGEMENTS ................................................................................................................ iii

ABSTRACT ..................................................................................................................................... iv

LIST OF ILLUSTRATIONS.............................................................................................................. ix

LIST OF TABLES ............................................................................................................................ xii

Chapter Page

1. INTRODUCTION............................................ 1

2. OBJECTIVES ................................................................................................................. 3

3. REVIEW OF LITERATURE ............................................................................................. 4

3.1 Nanograin Formation ....................................................................................... 5

3.1.1 SMAT Surface Mechanical Attrition Treatment ............................. 6

3.1.2 Physical Vapor Deposition ............................................................... 7

3.1.3 Electrodeposition .............................................................................. 7

3.1.4 ECAP Equal Channel Angular Pressing ....................................... 8

3.2 Wear Mechanisms ........................................................................................... 9

3.2.1 Fundamental Wear Mechanisms ..................................................... 9

3.2.1.1 Adhesive Wear ................................................................. 9

3.2.1.2 Abrasive Wear ................................................................ 10

3.2.2 Effects from Wear Debris ............................................................... 10

3.2.3 Aluminum and Titanium Wear Debris Visual Characterization ...... 12

3.3 Nanograin Formation During Wear ................................................................ 12

3.3.1 Mechanical Alloying ....................................................................... 14

3.3.1.1 Tribologically Transformed Structure ............................. 15

vi

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

7/90

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

8/90

viii

5.1 Microstructure................................................................................................. 40

5.2 Hardness ........................................................................................................ 43

5.2.1 Nanoindentation ............................................................................. 43

5.2.2 Knoop Hardness ............................................................................ 47

5.3 Surface Characteristics .................................................................................. 47

5.4 Tribological Behavior ...................................................................................... 48

5.4.1 Pure Nickel ..................................................................................... 50

5.4.2 Pure Aluminum ............................................................................... 58

5.4.3 Pure Titanium ................................................................................. 64

5.5 Tribological Behavior of Nanostructured Metals ............................................ 67

6. CONCLUSION .............................................................................................................. 69

REFERENCES ............................................................................................................................... 71

BIOGRAPHICAL INFORMATION .................................................................................................. 78

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

9/90

LIST OF ILLUSTRATIONS

Figure Page

3.1 Yield stress as a function of grain size [3]. ............................................................................... 5

3.2 One type of SMAT treatment - Ultra-sonic Shot Peening [8]. .................................................. 6

3.3 ECAP process: A rod is forced though a 90 bend to refine its structure [14]. ........................ 8

3.4 Aluminum frictional transients in different relative humidity:(a) 20%, (b) 46%, (c) 65% and (d) 95%. ............................................................................... 11

3.5 Wear track of cg material undergoes grain refinement during wear [22]. .............................. 13

3.6 Wear track of nc bulk material undergoes grain coarsening during wear [22]. ...................... 14

3.7 TTS layer in (a) maraging M250 and (b) Ti15V3Al3Cr3Sn [25]. .................................... 15

3.8 Comparison of the stress to operate a Frank-Read dislocation source andthe yield stress given by a Hall-Petch relation versus the grain size [5]................................ 16

3.9 Series of images extracted from a video of an in situ indentation on theultrafine-grained Al film: (a) the indenter is approaching the Al film, (b) bothgrains 1 and 4 grow at the expense of grain 2, (c) grain 2 has been eliminated

by the growth of its neighbors, and (d) grain 4 continues to grow, resulting ingrain 1 moving back to its original size and grain 4 consuming grain 3on its right side [27]. ............................................................................................................... 18

3.10 Dimpled fracture surface of Nickel-Tungsten tensile sample with an averagegrain size of 10 nm and with dimple diameters ranging from 100-300 nm [3]. .................... 19

3.11 Nanovoids forming ahead of a crack in nc nickel. In (d), the inset showsdislocations forming in front of the crack [32]. ...................................................................... 20

3.12 Tensile test result of nc nickel conducted at various strain rates [3]. ................................... 21

3.13 Tensile tests performed on various metals at various strain rates [3]. ................................. 21

3.14 Surface morphologies before and after fretting wear test of polycrystalline nickel;(a, d) 90 m grain size (b, e) 62 nm grain size (c, f) 13 nm grain size [33]. ........................ 23

3.15 Volume loss versus the inverse Vickers hardness [33]. ....................................................... 23

ix

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

10/90

3.16 Micrographs before wear tests: nickel coatings (a-3um, b-250 nm, c-16nm),cobalt coatings (d-2.5 um, e-220nm, f-18nm) and micrographs of samplesafter wear tests in the same order (nickel g-i, cobalt j-l) [34]. .............................................. 24

3.17 Micrograph of corroded surface of (a) nc and (b) cg copper [41]. ....................................... 26

3.18 Cyclic fatigue tests of nanocrystalline nickel (20-40 nm grains), ultra-finecrystalline nickel (300 nm grains), and microcrystalline bulk nickel [3]. ............................... 27

3.19 (a) Vickers hardness of multiple nickel C-2000 samples with varying SMATprocess time. (b) Fatigue life plot for the same samples [45]. ............................................. 28

3.20 Coefficient of friction of nickel with grain sizes of 8 nm, 22nm, and 61m [4]. ................... 28

3.21 Variation of fatigue crack growth rate with stress intensity factor in nickel [47]. .................. 29

3.22 Nickel samples subjected to sinusoidal fatigue loading at a frequency of 10 Hz and a load ratio of 0.3 [48]. ........................................................................................ 30

4.1 Container used to apply SMAT to samples. ........................................................................... 33

4.2 Loading profile for nanoindentation test. ................................................................................ 37

5.1 Optical micrographs of etched (a) Ni-1 (b) Al-2 and (c) Ti-1. ................................................. 40

5.2 -2 and low angle XRD results for (a) Ni-2 and (b) Al-3. ..................................................... 41

5.3 -2 scan of Ti-2 before (bulk SMAT Ti-2) and after (SMAT Layer Ti-2) theSMAT surface was removed. ................................................................................................. 42

5.4 Results of individual indentation tests compared to results of stepped tests

on SMAT Ni; (a) Reduced modulus and (b) Hardness. ......................................................... 445.5 Nanoindentation hardness results for bulk and SMAT materials. .......................................... 46

5.6 Reduced modulus results from bulk and SMAT materials. .................................................... 46

5.7 Typical 3D roughness profiles of materials ............................................................................ 48

5.8 COF as a function of sliding distance for bulk-Ni and SMAT-Ni. ........................................... 52

5.9 SEM micrographs showing typical wear track morphology in (a) Ni-1 withhigher magnification of wear track showing brittle nature within inset and(b) Ni-2 [scale bar is 500 m]. ............................................................................................... 53

5.10 EDS results from nickel wear tracks of Ni-1 and Ni-2. ......................................................... 54

5.11 COF as a function of sliding distance using a 6.35 mm diameter pin onSMAT nickel compared to testing with a 9.54 mm pin on SMAT and bulk nickel. ............... 54

5.12 Cross section of S-Ni directly after sputtering [scale bar is 2 m]. ...................................... 55

x

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

11/90

5.13 COF as a function of sliding distance for SMAT nickel, N-Ni, andsputtered deposited Ni. ........................................................................................................ 56

5.14 SEM micrographs showing typical wear track morphology in (a) N-Ni and(b) S-Ni [scale bar is 500 m]. ............................................................................................. 57

5.15 Wear debris from various nickel samples [scale bar is 10 m]. ........................................... 57

5.16 Typical 3D wear track profiles of nickel with various microstructures. ................................. 58

5.17 COF as a function of sliding distance for bulk aluminum (Al-1). .......................................... 59

5.18 (a) Al-1 Sample after wear tests with track 1 in the center and track 4 on the edge.Pin wear scar morphology for (b) track 1, (c) track 2, (d) track 3, and (e) track 4. .............. 60

5.19 EDS results from aluminum wear debris in different zones. ................................................ 60

5.20 COF as a function of sliding distance for bulk and SMAT aluminum. .................................. 61

5.21 SEM micrographs showing aluminum wear debris from (a) Al-1 track 1,(b) Al-1 track 2, (c) Al-1 track 3, (d) Al-1 track 4, (e) Al-3 track 1, and (f) Al-3 track 2. ........ 63

5.22 Typical 3D profile of wear tracks in different aluminum microstructures. ............................. 64

5.23 EDS results from titanium wear debris which displays consistent atomic composition. ...... 65

5.24 COF as a function of sliding wear of bulk and SMAT titanium. ............................................ 66

5.25 Typical SEM micrographs showing wear track morphology for (a) Ti-1 and(b) Ti-2 [scale bar is 500 m]. .............................................................................................. 66

5.26 Typical wear debris appearance in wear track of (a) Ti-1 and (b) Ti-2 ................................ 675.27 Typical 3D view of wear track section from (a) Ti-1 and (b) Ti-2. ........................................ 67

xi

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

12/90

xii

LIST OF TABLES

Table Page

4.1 Nominal composition of Nickel-200, wt%. .............................................................................. 31

4.2 Nominal composition of Aluminum-1100, wt%. ...................................................................... 31

4.3 Nominal composition of ASTM Grade 2 Titanium, wt%. ........................................................ 31

4.4 Test sample codes. ................................................................................................................ 32

4.5 SMAT application times. ........................................................................................................ 33

4.6 Magnetron sputtering conditions for sputtered nickel and titanium. ....................................... 34

4.7 Etchants used to reveal grain structure and average grain size of SMAT and bulk Ni, Al, and Ti. ................................................................................................ 35

5.1 Average grain size of Bulk, electrodeposited, and sputtered samples. ................................. 40

5.2 Peak location, FWHM, and approximate grain size of SMAT samplesfrom XRD analysis. ................................................................................................................ 42

5.3 Knoop hardness testing of N-Ni and S-Ni. ............................................................................. 47

5.4 Surface roughness of materials tested. .................................................................................. 48

5.5 Hertzian stress of materials studied. ...................................................................................... 49

5.6 Wear rates of materials tested. .............................................................................................. 50

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

13/90

CHAPTER 1

INTRODUCTION

Wear properties of materials play an important role in any material selection process.

Due to current material limitations, materials, such as titanium-base and aluminum-base alloys,

having high strength to weight ratio, cannot be selected for many applications with even a

moderate tribology requirement due to their poor wear resistance.

Pure titanium is of special interest in the medical industry due to its biocompatibility.

However, recent studies have shown that wear debris created by medical titanium alloys

creates irritation and inflammation around joints [1]. Ideally, pure titanium would be used, but

pure titanium easily galls and frets when in sliding or vibrating contact with itself or other metals

and conventional lubricants have trouble overcoming this [2]. Nickel has good tribological

properties and its alloys are commonly used. However, nickel is expensive, making cost a large

factor when selecting it for a particular wear application.

To overcome this, expensive alloys and complex surface coatings are continually being

developed for use with specific applications. However, little research has been conducted on

wear improvement through material processing. Processes such as electrodeposition and

severe plastic deformation can be applied to complex surfaces. Previous research has

indicated that material wear properties improve as the grain size at the surface drops below 100

nanometers [3]. Previous studies have shown that FCC nickel, with grain sizes below 100 nm,

has greater wear resistance than its counterpart with a micrometer grain size [4]. The smaller

grain size increases hardness and possibly prevents or restricts dislocation motion in the grain.

This leads to the possible improvements of other FCC metals, such as copper and aluminum,

and could also indicate an improvement in HCP metals, such as titanium. Surface mechanical

1

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

14/90

attrition treatment (SMAT) is a recent surface modification technique used to obtain a

nanocrystalline microstructure at the surface region of a material. The process is easy to apply

and produces a gradient grain size nanocrystalline microstructure. This is important because a

gradient grain size microstructure can produce a gradient in the hardness, resist crack

nucleation, and resist crack propagation. However, SMAT processing also creates a rough

surface, which could affect the wear behavior. This study focuses on the tribological effects of

the SMAT process on pure metals with FCC and HCP structures.

2

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

15/90

3

CHAPTER 2

OBJECTIVES

The present study is concerned with the processing-structure-property relationship in

nanostructured materials. FCC pure nickel, FCC pure aluminum, and HCP pure titanium were

chosen as model materials to study the wear behavior of a gradient nanocrystalline surface

structure. More specifically, the objectives are:

i) study the tribological behavior of gradient nanograined microstructures produced by

SMAT in three entirely different pure metals;

ii) to increase the scientific knowledge of the effects of surface processing

(nanocrystallization and hardening) in regards to tribological behavior.

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

16/90

CHAPTER 3

REVIEW OF LITERATURE

Optimal material properties are continually sought to improve processes and

applications. Nanocrystalline (nc) materials with grain sizes typically smaller than 100

nanometers have become the focus of attention due to their unique properties and deformation

mechanisms. These properties include high strength, strength and/or ductility dependence with

strain rate, increased resistance to tribological and environmentally-assisted damage, and their

potential for enhanced super-plastic deformation at lower temperatures with high strain rates [3].

The properties of nc materials have not been studied in great depth [3], however, there are

sufficient studies available to provide a general understanding of their potential benefits. Many

studies compare nc materials to ultra-fine-crystalline (ufc) materials with grain sizes ranging

from 100 nanometers to 1 micron, and ordinary microcrystalline materials (mc) with grain sizes

larger than 1 micron.

One of the benefits of nc materials is described by the Hall-Petch relationship, shown in

equation 1,

d

k yo y

+= (1)

where y is the yield stress, o is the stress required for dislocation motion (internal friction), ky

is the strengthening coefficient, and d is the grain size. This relation states that as the grain

size decreases, the strength of the material increases. This trend is valid for most grain sizes,however, studies have found materials in the lower end of the nc regime seldom follow the Hall-

Petch relationship [3]. Tests conducted on high quality specimens yield Hall-Petch plots similar

to Figure 3.1.

4

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

17/90

Figure 3.1 Yield stress as a function of grain size [3].It is believed that metallic materials with grain sizes smaller than 10 nm act more

closely like an amorphous metal, rather than a polycrystalline metal. It is generally accepted

that the reason for the change in the Hall-Petch slope, observed when grain sizes are below

100 nm, is suppression of inner grain dislocation mechanisms such as Frank-Read sources [5].

Other deformation methods, such as grain boundary sliding and nanovoids, are suspected to be

active in this small domain. Further discussion on these deformation activities will be provided

later in this chapter.

In the general nc range, small processing errors can have a great effect on the

materials properties. Material porosity is a common problem when forming nano-structured

materials and can be difficult to overcome. These issues are magnified as the desired grain

size decreases, especially below 10 nm. However, current techniques are beginning to create

reliable results for materials with grain sizes above 10 nm.

3.1 Nanograin Formation

Nanocrystalline materials can be created from three main types of processes; i) from

isolated nanometer sized particles, ii) from a noncrystalline structure, and iii) from a coarse

grained (cg) polycrystalline structure [6]. In the first process, nanometer sized particles are

created, typically by physical vapor deposition (PVD), chemical vapor deposition (CVD),

5

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

18/90

electrochemical and hydrothermal methods, and precipitation from a solution, and then

consolidated to form a bulk material where each particle becomes a nanocrystalline grain. The

second technique involves starting with an amorphous material and nucleating grains, possibly

by annealing or mechanical activation [7]. The third process involves increasing a

polycrystalline materials free energy by introducing numerous material defects and interfaces,

such as grain boundaries. This can be performed via ball milling, severe plastic deformation,

sliding wear, irradiation, and spark erosion [8]. Depending on the process used, different grain

refinement mechanisms transform the polycrystalline structure into a nanocrystalline structure.

Some common nanocrystalline processes are briefly described below.

3.1.1 SMAT Surface Mechanical Attrition Treatment

SMAT involves a grain refinement process with the application of mechanical

treatments [8]. Repeated localized severe plastic deformation can refine the grain size by the

creation of shear bands, small-angle grain boundaries, and completely random grain orientation

[8]. A common SMAT treatment is ultra-sonic shot peening (USSP). This process involves

shot, typically steel spheres, rapidly hitting the surface of a sample, as shown in Figure 3.2.

Figure 3.2 One type of SMAT treatment - Ultra-sonic Shot Peening [8]. [

This process has shown to produce a decrease in grain size in most materials, along

with an increase in hardness [6, 8-10]. The process is commonly used since its parameters are

easily controllable, ball-milling equipment can be utilized, and industry is familiar with the

process [6]. Another benefit of this process is the formation of a gradient microstructure,

6

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

19/90

approximately 50 m thick for most materials [11]. Depending on the material, the grains near

the surface are approximately 10 nm in size, and the grain size gradually increase until the bulk

material grain size is met. This creates a gradient microstructure and can have beneficial

applications, especially with tribology and fatigue.

3.1.2 Physical Vapor Deposition

Physical vapor deposition is a general name for many thin-film vacuum deposition

techniques. These processes typically involve plasma sputter bombardment or high

temperature vacuum evaporation. PVD is commonly used to create thin-films under 100 nm

thick, but can be used for thicker films. Different PVD processes include evaporative

deposition, sputter deposition, electron beam physical vapor deposition, cathodic arc deposition,

and pulsed laser deposition. Evaporative deposition heats the material to be deposited,

typically through resistive heating, under high vacuum, causing the material to evaporate and

deposit on the substrate positioned above it. Sputter deposition is performed in a controlled

ionized atmosphere, typically Argon. The ionized gas particles will be attracted to the positively

charged target, and upon impact, will knock an atom or cluster of atoms off of the surface. The

atoms will then deposit on the substrate.

Due to the complexity of these processes, many processing factors will affect the

deposited material structure, most notably temperature and pressure [12]. Generally, the

coatings created with this method are compact, and fully dense. However, porosity and film

adherence are two common problems.

3.1.3 Electrodeposition

Electrodeposition is an inexpensive process where a negative potential is applied to the

substrate material in an electrolytic bath. This process is easily controllable, and depending on

process parameters, the grain size of the resulting coating can also be controlled [13]. Complex

shapes can be electroplated and thicknesses of 100 m or more can be reached. The grain

size plated can be below 10 nm. Usually, to achieve grain sizes less than 20 nm, a grain refiner

7

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

20/90

must be utilized, such as Saccharin [3]. However, this can introduce impurities in the coating.

Along with this, it is very difficult to produce a fully dense electrodeposited coating. Nanovoids

can usually be found unless a compaction step is utilized [3]. Compaction will create a slightly

less dense material coating, which can have a significant effect on the properties of the

material.

3.1.4 ECAP Equal Channel Angular Pressing

Most processes producing nanocrystalline materials create only a thin surface layer.

However, the ECAP process can produce a bulk nc material. The ECAP process involves

forcing a cg or single crystal material through a die, typically with a 90 bend ( Figure 3.3). This

induces large strains in the material leading to grain refinement. A large benefit of the ECAP

process is the ability to apply a texture to the grains. One pass through an ECAP die will

produce elongated grains. If the material undergoes many ECAP passes, with rotation before

each pass, the microstructure can become homogenous [14]. One ECAP pass of a single

crystal material will produce an ufg polycrystalline material with a preferred texture [15].

Large Force

Figure 3.3 ECAP process: A rod is forced though a 90 bend to refine its structure [14].

8

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

21/90

3.2 Wear Mechanisms

Tribology classifies the type of wear that occurs by the relative motion between the

material surfaces in contact. These motions are typically referred to as sliding, rolling,

oscillating, impacting, and erosive wear. Tribological tests can also be described with other

variables based on factors such as the angle of action, lubricants, and environmental

conditions. Wear tests can also be performed with more than two contacting surfaces, such as

a ball bearing rolling between two plates and four-ball wear tests.

The hardness of two materials in contact generally indicates how the softer material will

wear. A material with a hardness ratio greater than 1.2 with respect to the counterface will tend

to undergo smooth sliding and less friction. A material with a hardness ratio less than 0.8 with

respect to the counterface will undergo rough sliding with higher friction coefficients [16].

3.2.1 Fundamental Wear Mechanisms

All basic sliding wear mechanisms fall into four categories: abrasive, adhesive, fatigue,

and oxidative wear [17]. Commonly, a wear system will not be a single type of wear, but a

combination of mechanisms. Adhesive and abrasive wear are the two dominant wear

mechanisms in this study, and are briefly described below.

3.2.1.1 Adhesive Wear

Other names are sometimes used to describe adhesive wear, such as scoring, galling,

and seizing, but they all describe the same process. As one surface slides relative to another,

surface asperities from the roughness of both materials contact, and create very high local

pressures. This localized high pressure creates plastic deformation and possibly welding

between the two surfaces. As the sliding continues, the welded region will typically break from

the softer material, creating a void on the surface and a protrusion on the second surface.

Debris is ultimately created when the welded asperities break, and adds to the wear of the

materials. The debris formation typically leads to abrasive wear.

9

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

22/90

Archards law can be used to describe the wear volume per unit sliding distance, W

[18],

H KPW = (2)

where P is the applied load, H is the hardness of the softer material, and K is the wear

coefficient. K is a relative comparison between the relative strength of the interface junction and

the strength of the asperities that make up the junction. A low value of K indicates that the

created junction material is softer than the surface asperities on both materials. Many factors

affect the wear coefficient, such as material hardness, roughness, and material type. Typically,

similar materials and very soft materials are the most susceptible to adhesive wear due to acold welding mechanism. Lubrication can greatly decrease the amount of adhesive wear

because of the thin film created that separates the two surfaces.

3.2.1.2 Abrasive Wear

Abrasive wear occurs when hard particles remove material from one or both of the

contacting surfaces. The hard particles can be embedded in one of the contacting surfaces,

such as in diamond saw blades, and is called two-body abrasive wear. The hard particles can

also be loose, between the two contacting surfaces. This process is called three-body abrasive

wear. The friction properties can differ greatly based on whether two-body wear or three-body

wear is present. In two-body wear, lubrication can actually increase the wear rate because it

squeezes out dull particles in the second material, and exposes a fresh sharp particle.

3.2.2 Effects from Wear Debris

Wear debris can have a strong effect on the wear process by turning a two-body wear

process into a three-body wear process. This will affect the materials lifetime. Studying the

composition and structure of the wear debris allows for a better understanding of what is

occurring during the wear process.

An excellent study example is aluminum. The tribological properties greatly depend on

the alloy content and the environment [19]. However, the mechanism of debris formation is the

10

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

23/90

same. Many researchers in the past assumed that wear debris created from aluminum

substrates is aluminum oxide [19]. This is assumed because the current wear theory predicts

that aluminum will be worn from the surface and oxidize due to the localized high temperatures.

However, this theory does not explain observations seen in reference [19]. Figure 3.4 shows

that there is a reduction in the friction coefficient followed by a rise before it reaches steady

state.

Figure 3.4 Aluminum frictional transients in different relative humidity: (a) 20%, (b) 46%, (c) 65%and (d) 95%.

The amount of humidity present during the tests affects how soon the dip is seen. As

the humidity increases, the dip occurs faster. This behavior cannot be simply attributed to the

formation of aluminum oxide but to the formation of both aluminum oxide and aluminum

hydroxide [19]. When a critical thickness of aluminum oxide is reached, a sudden drop in thefriction occurs, and wear debris begins to form. As the wear debris mechanically mixes with

moisture from the atmosphere, aluminum hydroxide is produced, and a rise in friction occurs.

This study indicates that mechanical mixing does not only occur between the two contacting

11

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

24/90

surfaces, but also with the atmosphere. This phenomenon can also be seen in steel with iron

hydroxides forming in high relative humidity environments, acting as a lubricant to reduce the

wear rate and decrease the friction coefficient [20].

3.2.3 Aluminum and Titanium Wear Debris Visual Characterization

Analyzing the wear debris from aluminum and titanium can be difficult because their

debris is commonly in an amorphous or nanocrystalline form. As was shown in the previous

section, aluminum wear debris forms aluminum oxide and then after mechanical mixing with

moisture from the atmosphere, forms amorphous aluminum hydroxide. Before careful analysis

is conducted, the debris can be preliminarily visually characterized by its color. Aluminum oxide

will appear grey to silver whereas aluminum hydroxides will appear black.

Titanium wear debris is typically titanium dioxide. Titanium dioxide exists in three

crystalline polymorph phases at room temperature; rutile (tetragonal crystal structure), anatase

(tetragonal crystal structure), and brookite (orthorhombic crystal structure). When pure titanium

oxidizes in humid air, the formation of rutile is thermodynamically favored [21]. Rutile appears

as a dark red to black color, anatase is an indigo to black color, and brookite is typically an

orange color. This basic knowledge can greatly help in the identification of the debris,

especially when only small quantities are created.

3.3 Nanograin Formation During Wear

When a cg material undergoes sliding wear, the large strain present at asperity contact

points will create a large amount of dislocations. As plastic strain grows in the bulk grains,

dislocations build up in the grain walls and the grain starts to shrink [16]. At the highest strains,

the grain structure will be refined with sharp grain boundary walls which resemble grains

produced at high temperature. These nano-sized grains will increase with depth until the bulk

grain size is obtained. As the sliding contact continues, the layer becomes thicker, but the

exponential decay remains the same [16]. Dislocations start to form at the interface of the

nano-structure surface layer. Cracks begin to grow at this interface, and as the cracks reach

12

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

25/90

the surface, large portions of material is removed creating large debris [22]. As debris formation

increases, the wear rate will also increase. Figure 3.5 shows a cross-section of a typical

steady-state wear track. The bulk material has coarse grains, whereas, the wear track region is

made up of grains in the nano-scale. In the transition zone between the nano-scaled grains and

the bulk coarse grained material, a large dislocation density exists.

Dislocation wallsnc surface grains

Coarse rains

Figure 3.5 Wear track of cg material undergoes grain refinement during wear [22].

The transferred material created during adhesive wear can be further deformed and

mechanically mixed with counterface material and/or environmental components. This results

with the transferred material possessing a very fine or nanostructured grain size [23].

Depending on the material, this can also lead to a thin-film layer forming on the wear region,

changing its wear characteristics. The thickness of this layer, typically called the tribologically

transformed structure (TTS), can be dependent on the size of the initial wear debris created,

ultimately affecting the wear characteristics. Mechanical alloying and TTS formation is

commonly used to describe the wear track and wear debris on coarse grained materials, and is

described in the following sub-sections.

If the grains at the surface are in the nano-scale regime before wear occurs,

dislocations will become difficult to form and less dislocation cracks will form at the surface.

13

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

26/90

The reduction in dislocations, especially near the surface, will prohibit nano-grains from

becoming loose debris. Friction at the surface will cause some grain growth, primarily by grain

boundary sliding and grain rotation, but also by temperature induced annealing. Figure 3.6

displays the resulting wear track morphology. This growth mechanism prevents debris

formation by suppressing dislocations, allowing for a decrease in the wear rate of the material.

It is the aim of this report to better understand the formation of this layer and its effects on

tribological properties.

Figure 3.6 Wear track of nc bulk material undergoes grain coarsening during wear [22].

3.3.1 Mechanical Alloying

A growing theory of Mechanical Alloying [24] has been used to explain how a

nanocrystalline surface layer is created on a substrate. Mechanical alloying is a process where

two materials are mixed together by a combination of high strain and locally high temperatures.

The alloy will have a large volume fraction of dislocations, resulting in the nucleation of nano-

sized grains in the debris or as a surface layer. This process can occur in the two-body or

three-body wear regimes. As a material is worn, material transfer can occur with little debris

formation, by diffusion or adhesion, creating a mixed layer at the surface of both counter

materials. In three-body wear, debris is formed, mixed with high strain rates, creating a new

third phase. Mechanical Alloying has important consequences in the wear of a material. The

14

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

27/90

surface layer, or debris, created can act as a lubricant or an abrasive, drastically changing the

tribological properties of the materials.

3.3.1.1 Tribologically Transformed Structure

As a material undergoes initial wear, in particular fretting wear (small oscillations,

usually smaller than the contact radius), a deformed region can form on the surface. This

region forms due to high strains between the two contacting surfaces and subsequent

recrystallization of the surface region. This new phase formed is the Tribologically Transformed

Structure (TTS) layer [25]. Figure 3.7 displays a TTS region in a maraging steel and a titanium

alloy. The sliding direction is horizontal. InFigure 3.7b, cracks can be seen running from the

TTS layer through the bulk material along the materials slip plane.

TTS layer

Figure 3.7 TTS layer in (a) maraging M250 and (b) Ti15V3Al3Cr3Sn [25].

The TTS layer starts to form as a thin layer on the substrate near the initial wear cycles.

As the wear cycles increase, the thickness increases until it reaches a steady state value.

Debris formation is minimal during this formation stage, however, once the critical thickness is

reached, debris will be created at the same formation rate as the TTS layer. The maximum

thickness of this layer will depend on the substrate and pin microstructures and the testingparameters. Due to high strains between the contacting surfaces, the TTS layer will have a

nanocrystalline microstructure, along with the debris that is created.

15

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

28/90

3.3.2 Deformation Mechanisms

To better understand wear properties of nc metallic materials, some of the proposed

deformation mechanisms need to be understood. Metallic materials with grains smaller than

100 nm suppress dislocation motion within the grains. As the grain size decreases, the stress

required to activate dislocation motion is higher than alternate deformation methods. Figure 3.8

displays the stress required to activate Frank-Read sources in copper and the flow stress as

calculated from the Hall-Petch relation. This trend suggests that dislocations emitted from the

grain are not the dominating deformation mode. Three proposed methods to explain the

deformation behavior include Grain Boundary Sliding, Grain Rotation, and Nanovoids.

Figure 3.8 Comparison of the stress to operate a Frank-Read dislocation source and the yieldstress given by a Hall-Petch relation versus the grain size [5].

3.3.2.1 Grain Boundary Sliding and Rotation

Grain boundary sliding and grain boundary rotation are usually tied to one another.

These deformation methods consist of the formation and annihilation of dislocations in the grain

boundaries. This process causes grains to grow and rotate as they deform.

To better understand the process, Ke et. al. [26] strained two samples of silver, one with

10 nm grains and the other with 110 nm grains. The results show no evidence of dislocations in

16

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

29/90

the sample with 10 nm grains and clear evidence of dislocation motion in the 110 nm grain

sample. The deformation in the 10 nm grain sample is believed to be caused by dislocation

motion within the grain boundaries.

The grain growth mechanism can be explained using thermodynamic principles. The

free energy of the system can be lowered by the growth of grains favorably oriented to respond

to the indentation load [27]. The grain reconfigurations are made possible, kinetically, by the

fact that the grain boundaries are very mobile even at room temperature [27]. This mobility may

result from the absorption of dislocations in the boundary, boundary sliding, or the simple

application of stress to the boundary [27]. This has been captured using TEM while deforming

an ufc aluminum film (Figure 3.9). Further tests performed by Jin et. al. [27] on iron did not

display the grain growth observed in aluminum. Nano-sized precipitates in the grain boundaries

pinned the grains preventing any sliding or rotating. Therefore, the composition and

microstructure of the nc metallic materials must be considered.

A further study by Vevecka et. al. [28] on the cyclic behavior of lead found that grain

boundary sliding occurred in low cycle fatigue, causing cracks to form around the grain

boundaries. These results were consistent with other results performed on lead and other

metallic materials.

Simulations conducted in the nc regime showed that the stress-strain response of the

nc metallic material exhibited an inelastic response at the grain boundaries, and an elastic

response in the interior of the grains [29]. Plastic deformation was rarely evident inside the

grains before failure. Instead, cracks would start to form as the grains reoriented, increasing the

internal strain until failure. The grain boundary cracking accounts for the decrease in

macroscopic ductility as a metallic materials grain size decreases.

17

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

30/90

Figure 3.9 Series of images extracted from a video of an in situ indentation on the ultrafine-grained Al film: (a) the indenter is approaching the Al film, (b) both grains 1 and 4 grow at the

expense of grain 2, (c) grain 2 has been eliminated by the growth of its neighbors, and (d) grain4 continues to grow, resulting in grain 1 moving back to its original size and grain 4 consuming

grain 3 on its right side [27].

3.3.2.2 Nanovoids

In nc metallic materials, the fracture surface of fatigued specimens contains spherical

dimples, usually a magnitude larger in diameter than the average grain size. The presence of

large dimples suggests that there is considerable local plastic deformation in the material as a

crack is progressing. Figure 3.10 shows a sample of nickel-tungsten with an average grain size

of 10 nm and spherical dimples of 100 to 300 nm in diameter [3].

18

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

31/90

Figure 3.10 Dimpled fracture surface of Nickel-Tungsten tensile sample with an average grainsize of 10 nm and with dimple diameters ranging from 100-300 nm [3].

[Tests have been performed to explain the dimpling of the surface. As a crack is

propagating through a nc metallic material, dislocations have been seen at the crack tip [3, 30-

32]. The dislocations emitted from the tip act to blunt the crack. They travel from the crack tip

to the grain boundaries. As the elastic energy grows in the grain boundaries, the triple junctions

will experience a high stress state. Grain boundary sliding can occur with the dislocation motion

within the boundary, and a small void will form at a triple junction. As the crack advances

closer, the void will grow until the crack reaches it. Nanovoids do not form at every triple

junction in front of a crack. They will only grow when the elastic energy at the triple junction

reaches a stress limit. This process can be seen in Figure 3.11, in a nc nickel. Figure 3.11(a)

shows three voids at three triple junctions in front of the crack tip. The crack grows into these

nanovoids. Figure 3.11(d) shows evidence of dislocation motion in front of the crack tip,

reinforcing previous simulations.

19

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

32/90

Figure 3.11 Nanovoids forming ahead of a crack in nc nickel. In (d), the inset showsdislocations forming in front of the crack [32].

3.4 Tribological Behavior of Nanostructured Metals

3.4.1 Material Behavior

As previously stated, the nc metallic materials display unique properties. Strain rate

dependent deformation is usually only observed in polymers and not in metals. However,

Figure 3.12 displays strain rate dependence of nc nickel with an average grain size of 40 nm.

This trend is not only seen in nc nickel, but in most nc metallic materials. Figure 3.13

displays results from a variety of tests performed on mc (black data points), ufc (blue data

points), and nc (red data points) metallic materials. Figure 3.13 shows very little change in yield

strength at low strain rates for mc and ufc metallic materials but shows a strain rate dependence

for the nc materials. However, at very high strain rates above 1000 s -1, there is a dependence

20

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

33/90

on strain rate for all materials. When a load is applied extremely fast to any material, it will

prevent dislocation motion and delay failure. Nanocrystalline materials already suppress

dislocations due to their small grain size, and therefore display a strain rate dependence. This

strain rate dependence could have various implications to structural applications, where low

strain rates, such as below 0.1 s -1, are observed.

Figure 3.12 Tensile test result of nc nickel conducted at various strain rates [3].

Figure 3.13 Tensile tests performed on various metals at various strain rates [3].

21

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

34/90

3.4.1.1 Wear Resistance

Sliding wear in mc metallic materials creates large plastic strains in the grains at the

surface [24]. Strain at the surface of the material causes grain refinement with dislocation

mechanisms, creating a rough surface and wear debris. This leads to cracks forming at the

surface and reducing the metallic materials life. Since inner granular dislocation motion is

suppressed with nc metallic materials, the wear rate and crack formation decrease as compared

to mc metallic materials. Results from wear tests performed by Jeong et. al. [33] on mc nickel

and electroplated nc nickel are shown in Figure 3.14.

It must be noted that the surface topography is not consistent between the 90 m nickel

(Figure 3.14 (a, d)) and the 13 nm nickel ( Figure 3.14 (c, f)), however, the fretting observed

along the wear track is a good indication of the wear resistance. Comparing (d), (e), and (f),

deep gouges along with fretting can be seen in the mc sample (a, d). As the grain size

becomes smaller, as in (f), the wear track becomes shallow and minimal fretting is observed.

This correlates well with the Hall-Petch relationship: as the grain size decreases, the strength

and hardness increases. The decrease in the gouging effect will also have a positive impact on

the fatigue life of the material by preventing crack nucleation sites. This test also displayed a

decrease in the friction coefficient from the mc nickel to the nc nickel by over 50%. The

increase in wear resistance can also be seen in Figure 3.15, where the volume loss of each

sample was plotted against the Vickers hardness.

22

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

35/90

Figure 3.14 Surface morphologies before and after fretting wear test of polycrystalline nickel; (a,d) 90 m grain size (b, e) 62 nm grain size (c, f) 13 nm grain size [33].

Figure 3.15 Volume loss versus the inverse Vickers hardness [33].

A similar test was performed by Wang et. al. [34]. A 50 m thick nickel and cobalt layer

was electrodeposited on a substrate. Figure 3.16 displays the electroplated nickel before pin-

on-disk wear tests (a-c) and after the wear tests (g-i) as well as the electroplated cobalt before

wear tests (d-f) and after wear tests (j-l). As in Figure 3.14, the wear resistance of the nickel

23

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

36/90

samples increases as the grain size decreases. The material becomes harder and the

mechanisms of deformation are suppressed as the grain size decreases, resulting in the

decrease in wear rate. The cobalt samples also display a slight decrease in wear rate with

decreasing grain size, but at a much smaller scale. The wear rate for both the mc and nc

cobalt coatings are far lower than the nickel coatings. This can be attributed to better resistance

of HCP structure to adhesion interactions compared to other atomic structures [34].

(g)

(h)

(i)

(j)

(k)

(l)

Figure 3.16 Micrographs before wear tests: nickel coatings (a-3um, b-250 nm, c-16nm), cobaltcoatings (d-2.5 um, e-220nm, f-18nm) and micrographs of samples after wear tests in the same

order (nickel g-i, cobalt j-l) [34].

Only a handful of studies have explored the grain growth phenomenon in

nanocrystalline materials subjected to wear. Nanocrystalline aluminum has shown a decrease

in wear debris compared to mc aluminum at high loads and high sliding velocities [35].

However, large plastic strains exceeding 10 (1000%) are known to be present at the interface[16], and the decrease in wear debris can partially be attributed to its increased shear strength

with decreasing grain size.

24

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

37/90

Ultra-fine grained titanium, created using an ECAP technique, showed improved friction

and lower wear rates compared to an annealed coarse grained titanium. Titanium oxide was

produced on both the cg and ufg titanium specimens. It is also important to note that the wear

mechanisms tended to form micro-plowing and delamination of the surface layer. This suggests

that high contact stresses are present just below the contact surface, and fatigue affected both

ufg and cg titanium. More titanium oxide was found on the cg wear track when compared to the

ufg wear track. Even though their friction coefficients were similar, the ufg had a lower wear

rate [36].

3.4.1.2 Corrosion Resistance

A possible added benefit of a nanocrystalline surface layer is an improvement in a

materials corrosion resistance. A consensus has not been made on whether the properties

improve, degrade, or lie somewhere between. The corrosion resistance of electroplated nickel

has been found to improve with deceasing grain size [37]. This is due to the ability of the

surface to rapidly form a passivating layer of nickel hydroxide since the surface energy is high

from the exposed grain boundaries.

In titanium, Balyanov et. al. [38] concluded that ufg titanium (approximately 300 nm)

improved the corrosion resistance by rapidly creating a passivating layer, due to the higher

surface energy. However, Garbacz et. al. [39] showed that nc titanium had little to no effect on

the corrosion resistance, and actually degraded the corrosion resistance. [40][41]

In copper, generally it has been found that the corrosion resistance decreases with nc-

copper [40,41]. However, voids forming during the processing of the nc copper were suspected

as one reason for the decrease in the corrosion resistance [41].

Finally, Zhang et. al. [42] concluded that nc zirconium exhibited better corrosion

resistance compared to mc zirconium, Mordyuk et. al. [43] reported that ultrasonic shot peened

austenitic steel showed improved corrosion resistance by increasing the volume fraction of

25

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

38/90

strain-induced martensite, and Aghion et. al. [44] concluded that the corrosion resistance of nc

magnesium is lower than that of the mc magnesium.

All of the aforementioned reports agree on two points: the large area fraction of grain

boundaries present at the surface of nc-materials allow the material to corrode more uniformly

and pitting is more prevalent in cg-materials since impurities tend to segregate to grain

boundaries. Figure 3.17 shows the surface morphology of nc copper and cg copper. The nc

copper had a slightly higher corrosion rate, however, as can be seen in the micrograph, the

surface corroded much more uniformly.

Figure 3.17 Micrograph of corroded surface of (a) nc and (b) cg copper [41].

3.4.2 Fatigue Response

Nanocrystalline metallic materials show also an improvement in the cyclic response.

Fatigue tests shown in Figure 3.18 were conducted at a R-value of zero and a frequency of 1

Hz, where R is the applied stress ratio of minimum to maximum stress. The results show an

increase in the high cycle fatigue strength of nc nickel. However, further experiments have also

shown that metallic ufc materials undergo cyclic softening and a weakening of their low-cyclefatigue resistance [3]. Coarse grained metallic materials tend to resist low cycle fatigue better

than nc grained materials but are less resistant to high-cycle fatigue.

26

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

39/90

Figure 3.18 Cyclic fatigue tests of nanocrystalline nickel (20-40 nm grains), ultra-fine crystalline

nickel (300 nm grains), and microcrystalline bulk nickel [3].Many studies have shown that SMAT introduces plastic deformation in the surface of a

material creating a nc zone and producing compressive stresses near the surface. A grain size

distribution will be created where nc grains will be developed at the surface followed by a grain

size increase with distance from the surface [45]. This can increase the life of the material by

suppressing the tensile zone in front of any stress riser, preventing cracks from forming.

However, it must be noted that applying a SMAT process for too long will provide no benefit and

could degrade the properties of the material as seen in Figure 3.19 (b). The SMAT application

applied for 30 minutes has a much higher fatigue limit than the original or other SMAT

processes. Figure 3.19a shows that there is little difference in hardness values for the varied

times. The properties will degrade because dislocations, microcracks, and surface

contaminates will increase as the process time increases creating nucleation points for cracks.

In stainless steels, annealing can be conducted after the SMAT process to relax the

compressive stress while maintaining the nc zone at the surface. This will lead to improved

fatigue life through a combination of strain hardening and strain-induced martensitic

transformation [46].

27

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

40/90

(a) (b)

Figure 3.19 (a) Vickers hardness of multiple nickel C-2000 samples with varying SMAT processtime. (b) Fatigue life plot for the same samples [45].

Figure 3.20 displays the improved coefficient of friction (COF) for nickel as the grain

size decreases. The lower COF produces less wear and hence less gouging will be present,

eliminating crack nucleation sites.

Once a crack forms, it will inevitably grow. The crack growth rate in nc metallic

materials is larger than the growth rate of mc metallic materials ( Figure 3.21). In this graph, the

nc material shifts the curve to the left, similar to increasing the load ratio (R). The load ratio is

the minimum stress divided by the maximum stress. As the ratio is increased, a larger is

applied, and the crack propagation rate will increase.

Figure 3.20 Coefficient of friction of nickel with grain sizes of 8 nm, 22nm, and 61m [4].

28

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

41/90

Figure 3.21 Variation of fatigue crack growth rate with stress intensity factor in nickel [47].

Multiple factors exist to explain the increase in crack growth. First, as stated previously,

grain boundary rotation, sliding, and nanovoids become the dominate deformation process in a

nc metallic material. This allows nanovoids to form ahead of the crack, creating a path for the

crack to follow. Second, with small grains, the active slip planes in each nano-sized grain will

only slightly deflect the crack. This enables a crack to propagate in a relative straight path,

mimicking a brittle fracture. This reduces any crack closure mechanisms that are present in cg

materials. Figure 3.22 compares cracks in mc, ufc, and nc nickel. As a crack propagates along

large granular slip planes, it will create a saw tooth type path, as seen in Figure 3.22a. When a

material is stressed, the crack faces can come in contact with each other, reducing the stress at

the crack tip, and slowing the crack growth rate. However, with nc metallic materials, the grains

are small enough to keep the crack deflection minimal, keeping the crack strictly in mode I

fracture. Simulations of crack growth in nanocrystalline metallic materials have also been

conducted [48]. The models predict that as the grain size becomes smaller, the rate of crack

growth will increase.

29

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

42/90

Figure 3.22 Nickel samples subjected to sinusoidal fatigue loading at a frequency of 10 Hz and

a load ratio of 0.3 [48].

Since the majority of fatigue cracks initiate at the surface, an ideal material will have a

nc surface layer with a gradient microstructure, with the grain size increasing with the distance

from the surface, similar to the microstructure created by the SMAT process.

30

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

43/90

CHAPTER 4

EXPERIMENTAL

4.1 Materials

The selected materials used were nickel-200, Aluminum-1100, and ASTM (American

Society for Testing and Materials) Grade 2 Titanium. All bulk model metals were selected

because their purity is greater than 99% and they are commonly used in tribological studies.

The material purity allows for accurate observation of the tribological behavior of their single

element form. The nominal composition of the bulk materials is shown in the following tables

[49].

Table 4.1 Nominal composition of Nickel-200, wt%.

Ni C Mn Fe Si Cu99.5 0.06 0.25 0.15 0.05 0.05

Table 4.2 Nominal composition of Aluminum-1100, wt%.

Al Si plus Fe Cu Mn Zn Other 99.0 min 0.95 0.05 to 0.20 0.05 0.10 0.15

Table 4.3 Nominal composition of ASTM Grade 2 Titanium, wt%.

Ti C H O N Fe99.3 min 0.10 0.0125 0.25 0.03 0.30

The bulk Nickel-200 and Aluminum-1100 was ordered in the form of 38.1 mm diameter

bar stock. The Grade 2 Titanium was ordered in the form of 51.5 cm diameter bar stock. Discs

38 mm in diameter disks and 1.6 mm thick were cut on a lathe for preparing bulk and SMAT

samples.

31

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

44/90

A large number of samples have been created and labeled for easier identification.

Table 4.4 displays the sample name, material, and the way it was made and/or treated. The

following sections will detail the way the samples were prepared and treated.

Table 4.4 Test sample codes.

Code Material DescriptionNi-1 Nickel Coarse grain disk from bar-stockNi-2 Nickel SMATN-Ni Nickel ElectrodepositedS-Ni Nickel Magnetron Sputtered Al-1 Aluminum Coarse grain disk from bar-stock Al-2 Aluminum Coarse grain disk from bar-stock Al-3 Aluminum SMAT

T-Al Aluminum Thermally evaporatedTi-1 Titanium Coarse grain disk from bar-stockTi-2 Titanium SMATS-Ti Titanium Magnetron Sputtered

4.1.1 Bulk Specimens

The nickel, aluminum, and titanium bulk samples were ground with silicon carbide

papers down to 800 grit to remove any machining marks and layers with residual stresses.

They were then progressively polished down to 0.25 m diamond paste with methanol. This

was the final polish step for aluminum. Polishing with 0.05 m alumina paste and methanol wasthe final step for nickel and titanium samples. All samples were cleaned with methanol in an

ultrasonic cleaner for 15 minutes after their final polish.

4.1.2 SMAT Specimens

The SMAT samples were polished and cleaned following the same steps as the bulk

samples. To apply SMAT to the sample discs, a Spex 8000M Mixer/Mill was utilized. A Spex

8001 hardened steel vial, typically used to ball mill powders, was modified to hold the sample

discs. Four 3.18 mm diameter 440 steel balls were used for the treatment at a frequency of 18

Hz.

32

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

45/90

Cap

Backing Plate

Retention Ring

Sample

Container

Ball

Figure 4.1 Container used to apply SMAT to samples.[

The treatment times differed for the different materials. Many SMAT treatment times

can be found in the literature for aluminum [9, 22, 50, 51] and titanium [10,11] and the SMAT

application time for the present work is based on those findings. However, very little research

has studied the optimal SMAT application times for nickel. Therefore, the nickel SMAT

application time was based on work performed by Chun Sheng-Wen et. al. [52], and can be

verified by a recent publication studying different SMAT application times on nickel [53]. The

SMAT times used in the present work are shown in Table 4.5 .

Table 4.5 SMAT application times.

Sample Code Time (min)Nickel-200 Ni-2 60

Aluminum-1100 Al-3 15Titanium Grade 2 Ti-2 60

4.1.3 Electrodeposited Nickel

An electrodeposited nickel sample was also purchased for comparison purposes. The

electrodeposition conditions are described elsewhere [54]. The sample was deposited with a

current density of 15 mA/cm2 at a temperature of 30C. Saccharin, Coumarin, and Lauryl

sulfate were used in the bath composition to help create a more dense material. The resulting

33

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

46/90

material is a fully dense nanocrystalline nickel sheet 1.5 mm thick. The sample is labeled N-Ni.

It was polished similarly as the bulk and SMAT nickel.

4.1.4 Thin-film Samples

To help evaluate the benefits of a nanocrystalline surface layer in tribological

applications, thick-filmed samples were also prepared. The nickel and titanium samples were

prepared using a magnetron sputtering technique, and the aluminum sample was prepared by

thermal evaporation. Each sample is approximately 1 m thick.

4.1.4.1 Magnetron Sputtering

Following well-known and established techniques, particularly developed by Thorton

[12], multiple nickel and titanium samples were prepared to determine the optimal coating

parameters using the magnetron sputtering equipment in the Surface and Nano Engineering

Laboratory. Silicon wafers with a {111} orientation were used as a substrate. Each substrate

was RF sputter cleaned by applying a 35 W bias to the substrate for 15 minutes in an argon

atmosphere of 6 mtorr. This removed any oxide layer from the substrate surface to provide for

a better bonding surface. Also, each target was cleaned by pre-sputtering while the shutter

remained above the target, covering the sample. The sputtering deposition for both samples

was conducted with 60 W of DC power applied to the target and 40 W of RF power applied to

the substrate. The nickel was sputtered directly on silicon substrate, whereas the titanium was

sputtered on a thin layer of chromium, which acted as a bond layer. The deposition conditions

for the latter two polycrystalline samples are shown below in Table 4.6 .

Table 4.6 Magnetron sputtering conditions for sputtered nickel and titanium.

SampleBase

Pressure

(torr)

SubstrateTemperature

(C)

Rotation

(RPM)

Time

(min)

Thickness

(m)S-Ni 5.9x10-7 400 10 165 1.00S-Ti 4.8x10-7 550 10 360 0.90

34

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

47/90

4.1.4.2 Thermal Evaporation

Magnetron sputtering of aluminum thin films can result in a porous and columnar

structured film. Therefore, thermal evaporation was utilized. Following Semaltianos [55], a high

deposition rate was used to create a thick aluminum coating on a silicon wafer substrate. A

Varian NRC 3117 thermal evaporator was used and the aluminum was deposited at the

evaporators maximum deposition rate of 4 /s and a pressure of 10 -5 torr.

4.2 Microstructure

4.2.1 Grain Size

The grain size of the bulk material was determined by following the Abrams Three-

Circle Procedure described in ASTM E112-96. Table 4.7 displays the etchant used on each

material.

Table 4.7 Etchants used to reveal grain structure and average grain size of SMAT and bulk Ni, Al, and Ti.

Material ASTM E407-07Etchant Composition

Nickel 13350 mL HNO3 50 mL Acetic

Acid

Aluminum 2 3 mL HF100 mL water

Titanium 18710 mL HF30 mL HNO3 50 mL water

The approximate grain size of N-Ni and the sputter deposited samples was determined

by x-ray diffraction (XRD) using a Siemens D-500 Diffractometer with Cu K radiation. -2

scans were performed from 20 to 80 with a step size of 0.008 and a hold time of 5 s. The

approximate grain size was obtained using the Debye-Scherrer formula by measuring the full

width at half maximum (FWHM) of the {111} peak in N-Ni, S-Ni, and T-Al and measuring the

1010peak in titanium.The SMAT samples have a gradient microstructure, and therefore the obtained grain

size from XRD represents the average size in the probed volume. The SMAT nickel and

35

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

48/90

aluminum were analyzed by comparing -2 scans of their {111} peak with a fixed 12 incident

angle 2 scan of the same peak. The step size was 0.008 with a hold time of five seconds.

The SMAT titanium was cut into quarters using an Allied Techcut 5 diamond saw, with a

water-based coolant, at 800 RPM and a cutting speed of 0.5 cm/min. One quarter was then

mounted in epoxy. Finally, the SMAT sample was sectioned with a diamond saw at the same

conditions, at a slight angle with respect to the SMAT surface. The sectioning created a thin

SMAT layer sample and a thin bulk layer from the back of the SMAT surface. Due to the angle

of the cut, the thickness of the sample varied approximately from 250 m to 90 m. -2 scans

were performed on both slices, with the SMAT slice positioned such that the x-rays penetrated

near the 90 m edge. The grain size was approximated using a the Debye-Scherrer formula.

4.3 Material Characterization

4.3.1 Nanoindentation

A Hysitron TI 700 Ubi1 nanomechanical test instrument was used to characterize the

hardness of the samples. A Berkovich indenter tip was used to perform the measurements.

The measurements characterized the hardness of the samples versus the depth from the

surface. A single indent profile was used, and consists of a constant loading rate of 50 nm/s, a

1 second hold, followed by a 25 nm/s unload. Figure 4.2 displays the loading profile graphically.

The Hysitron software package, Tribosoft, analyzed the resulting load versus displacement

data, and using a power law equation, calculated the hardness and reduced modulus. At least

four indents were performed at any one indent depth and the average is reported.

36

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

49/90

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

50/90

are moving average trend line fits, using 100 COF data points. This is used to remove chatter

of the graphs so that they can be easily read and compared.

The wear volume for each test was calculated by averaging four 90-separated cross-

sections of the wear track and multiplying it by the circumference of the track. The wear rate

was then calculated by dividing the wear volume by the contact load and sliding distance. The

wear rate of the alumina pin was calculated by measuring the wear scar diameter. Two

measurements were taken, 90 apart and then averaged to obtain the scar diameter. Following

ASTM G99-05, the wear pin volume was then calculated by equation 3.

r

d W P

64

4

= (3)

where WP = Wear Volume, d = average wear scar diameter, and r = wear pin radius.

4.5 Hertzian Stress

The initial stress on the sample surface from the loaded pin must be known to

understand the wear test starting conditions. High Hertzian stresses can cause yielding of the

surface material, which can affect the outcome of the test. The maximum contact stress

between the pin and the surface is calculated by finding the maximum contact area between the

pin and the surface. This is obtained by following a well established formula as presented by

Shigley [56]. The radius of the contact area is calculated using equation 4.

(4)F is the applied force, v is Poissons ratio, E is the modulus of elasticity, and d is the diameter of

the contacting surfaces. The maximum contac tress then is calculated by equation 5.t s

p F (5)

The maximum deflection, maximum shear stress, and location of maximum shear stress can

then be calculated as presented in Stachowiak and Batchelor [57].

38

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

51/90

The SMAT materials have a rough surface, and therefore the contact cannot be

approximated as a sphere on a smooth surface. The SMAT surface topography is

characterized by many dimples and peaks, or asperities. The maximum Hertzian contact stress

occurs when the pin is positioned on one of these asperities. The Hertzian stress can then be

modeled as a sphere on an infinitely long cylinder. Therefore, the Hertzian contact stress was

calculated using the approximate contact stress calculations for elliptical contact areas

developed by Hamrock and Dowson, and as described in Stachowiak and Batchelor [57].

4.6 Scanning Electron Microscopy (SEM)

A Hitachi S-3000N variable pressure scanning electron microscope equipped with

energy-dispersive spectrometry (EDS) was utilized to analyze the wear track, wear debris, and

quality of deposited films. An electron beam energy of 10 to 25 kV was used, depending on the

material being analyzed. All samples were mounted on an aluminum stub with graphite tape.

After the tests, wear debris was analyzed by placing the debris on copper tape, and then

mounting the copper tape to an aluminum mount with graphite tape.

4.7 Surface Profilometry

The volume loss of the wear tracks was measured using a Veeco Wyco NT9100 optical

profilometer. Each sample was scanned using the units VSI (Vertical Scanning interferometry)

mode at various magnifications and scan lengths. The surface roughness of each sample was

calculated by using the accompanying Veeco software, Vision. Four measurements were taken

at 5.0X magnification, which corresponds to a surface area of 1.2 mm 2, and the results were

averaged.

39

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

52/90

CHAPTER 5

RESULTS AND DISCUSSION

5.1 Microstructure

The approximate grain size of the SMAT samples and the PVD aluminum and titanium

samples was calculated by using x-ray diffraction. The results for the bulk, electrodeposited,

and sputtered samples are displayed in Table 5.1 .

S-Ni displays a XRD pattern similar to N-Ni, indicating that the sputtered film has a

grain size in the nanoscale. S-Ti displays many oxide peaks along with the titanium peaks,

which indicates that the sputtered titanium is possibly not fully dense and is oxidized. T-Al has

a large grain size and an oxide layer is clearly present on the surface, along with XRD oxide

peaks. This indicates that S-Ti and T-Al sputtered films are not high quality films.

(c)(b)(a)

Figure 5.1 Optical micrographs of etched (a) Ni-1 (b) Al-2 and (c) Ti-1.

Table 5.1 Average grain size of Bulk, electrodeposited, and sputtered samples.

Sample Grain SizeNi-1 39.2 3.1 mN-Ni 54.8 nmS-Ni 75.4 nm Al-2 291.3 30.7 mT-Al >200 nmTi-1 34.0 3.8 mS-Ti 126 nm

[

40

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

53/90

Direct evidence of a gradient microstructure in the SMAT materials is difficult to obtain.

Multiple studies indicate that SMAT produces a gradient microstructure [9-11, 22, 50-53]. XRD

analysis can indicate that a gradient does exist. Peak broadening is a leading indication of

grain refinement. A problem with regular -2 diffraction is the significant depth of penetration

inhibits surface probing. With polycrystalline materials, the incident x-ray beam probes a large

volume and Braggs law will always be satisfied due to the presence of many randomly oriented

grains. Therefore, to avoid penetrating deep in the surface, a diffraction scan centered at 12

with the surface was conducted for the SMAT nickel and SMAT aluminum samples. The results

are shown in Figure 5.2 . Titanium has a higher x-ray absorption coefficient than nickel and

aluminum, and a 12 scan provided only a weak peak. Therefore, the SMAT titanium was

mounted in epoxy and sliced along the SMAT surface, creating a 100 m thin SMAT foil and a 1

mm thick bulk foil. -2 scans were then performed on both slices, and the results were

compared. This method provides an insight on the top layer of Ti-2 and the bulk titanium in Ti-

2. Table 5.2 displays the results of the XRD analysis.

(a) (b)

Figure 5.2 -2 and low angle XRD results for (a) Ni-2 and (b) Al-3.

41

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

54/90

Figure 5.3 -2 scan of Ti-2 before (bulk SMAT Ti-2) and after (SMAT Layer Ti-2) the SMATsurface was removed.

Table 5.2 Peak location, FWHM, and approximate grain size of SMAT samples from XRDanalysis.

Scan Type 2

(deg.) FWHM (deg.)

Grain Size (nm)

Ni-2{111}

-2 Scan 44.43 0.25 34.9 = 12 44.39 0.34 25.2

Al-3{111}

-2 Scan 38.42 0.18 47.8 = 12 38.39 0.24 34.5

Ti-2{ }

-2 SMAT Edge 39.94 0.50 33.0 -2 Bulk 40.25 0.26 16.9

The XRD scans on the SMAT nickel, aluminum, and titanium display both peak

broadening and peak shifting. The peak shifting of the SMAT nickel and aluminum are similar in

both scans, indicating that a residual stress is present in the material. This residual stress is

likely from the SMAT processing. It is well known that residual stresses will shift XRD peak

positions. A shift to a lower Bragg angle indicates a tensile stress is present. This at first

seems unlikely, however, the SMAT samples were prepared in a vial, with the edges of the

round sample fully clamped. As the SMAT process is applied, the material will deform, and

since it is clamped, will cause the sample to bow out, creating a tensile stress near the SMAT

surface and a compressive stress on the back surface. This can be seen in the SMAT titanium.

The SMAT layer displays a shift lower than Braggs angle, and the back bulk part of Ti-2

displays a shift higher than Braggs angle, indicating a compressive stress. This is not direct

42

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

55/90

evidence of a gradient microstructure, however, by the mechanics of the SMAT process, it is not

likely that a thin layer of smaller grains forms at the surface, with no grain size gradient to form.

5.2 Hardness

5.2.1 Nanoindentation

Since six indentations were performed at the same location but at different depths, work

hardening is a concern. An existing SMAT nickel sample was used as a test specimen. The

sample was processed for one hour with six 3 mm diameter harden steel balls. Eight individual

indents were performed at each displacement of the sample, followed by eight indents using the

multi-displacement loading curve. The results, displayed in Figure 5.4, show that there was

only a small difference between the two tests. The difference in hardness can largely be

attributed to the loading rate used for the tests. The individual indents were performed with a

set loading, hold, and unloading time of five seconds, whereas the step indents were performed

using the loading profile shown in Figure 4.2. There is a larger error with the step indents and a

slight decrease in the hardness is observed. However, this is within error limits and therefore

can be used.

43

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

56/90

(a)

(b)

Figure 5.4 Results of individual indentation tests compared to results of stepped tests on SMATNi; (a) Reduced modulus and (b) Hardness.

44

7/28/2019 Tribological Behavior of Nanocrystalline Metals Prepared by Surface Mechanical Xxx

57/90

The mc and SMAT materials were tested for hardness. The results are shown in Figure

5.5 and Figure 5.6. The results show that the SMAT nickel hardness is lower than the mc-nickel

and the SMAT aluminum and SMAT titanium are showing hardness values only slightly higher

than their mc counterpart. When the reduced modulus is examined, the SMAT nickel modulus

is drastically lower than the mc nickel, and deviations are seen in the SMAT aluminum and

SMAT titanium. The change in reduced modulus is a sign that the rough SMAT surface may

have affected the values